`CVE-2026-2587`

9.6 Critical No exploit published

Description

A critical Remote Code Execution (RCE) vulnerability was identified in the server-side template rendering mechanism used by the Glassfish gadget handler. The application processes .xml files and evaluates user-supplied values within a context where Expression Language (EL) “expressions” are processed without proper sanitization or escaping. By injecting expressions such as #{7*7}, the server returns 49, confirming server-side EL evaluation. This issue allows a remote attacker to fully compromise the underlying host, enabling capabilities as reading/modifying data, executing arbitrary commands, persistence, and lateral movement. This issue affects Eclipse GlassFish: from 8.0.0 to 8.0.1, fixed in 8.0.2; 7.1.0, fixed in 7.1.1; from 7.0.0 to 7.0.25, fixed in 7.0.26. Impact on versions from 5.1.0 to 6.2.5 is unknown.

CWE-917

Exploitation Status

No public exploit

Nothing published shows this CVE being exploited. The 1 repository below are public repositories under the 5-star bar: published, but with nobody other than their authors having looked.

CVSS E:U

CISA KEV not listed
CISA Vulnrichment exploitation: none
Indexed PoC 1 indexed Published artifacts you can open, in GitHub PoC; first seen 2026-05-20.
EPSS 0.65% chance in 30 days

1 of these are GitHub repositories below the 5★ evidence bar. They are listed because a person may still want to open one; they do not count toward the exploit maturity above.

Index	Artifact	Stars	First seen
GitHub PoC	CVE-2026-2587 PoC validator for Eclipse GlassFish EL Injection RCE in the admin console gadget.jsf handler. Safe authenticated vulnerability scanner for authorized testing.	1	2026-05-20

Index

Artifact

Stars

First seen

GitHub PoC

CVE-2026-2587 PoC validator for Eclipse GlassFish EL Injection RCE in the admin console gadget.jsf handler. Safe authenticated vulnerability scanner for authorized testing.

2026-05-20

Waratek Defense Posture

No exploit published

In RASP's scope with an upstream fix path, but nobody has published a working exploit. That is a statement about what has been published, not a finding that the defect is unreachable or unimportant.

Protection none
Action unverified patch rule
Review automated upstream-data
Record active

Decided by no-exploit-published : A published working exploit for this candidate, not only a proof of concept, a forecast, or nothing at all

Finding This CVE is not listed in CISA KEV, CISA Vulnrichment records no active exploitation, and no reviewed exploit catalogue carries an entry for it. Nobody has published something a stranger can run, so it is not a candidate for a rule ahead of the ones where somebody has; an upstream fix path does exist, so the remediation half of the candidacy still holds. This is a statement about what has been published, not about whether the defect is reachable or serious.

Decided by	`no-exploit-published` : A published working exploit for this candidate, not only a proof of concept, a forecast, or nothing at all
Finding	This CVE is not listed in CISA KEV, CISA Vulnrichment records no active exploitation, and no reviewed exploit catalogue carries an entry for it. Nobody has published something a stranger can run, so it is not a candidate for a rule ahead of the ones where somebody has; an upstream fix path does exist, so the remediation half of the candidacy still holds. This is a statement about what has been published, not about whether the defect is reachable or serious.

CVSS

9.6 CRITICAL v3.1 CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H 1 metric

VRT selects the newest version's highest entry and publishes it as cvssScore, newest rather than largest because scores are not comparable across versions, and highest rather than first because the first entry is frequently a CNA placeholder scoring 0.0 over NVD's own analysis.

Version Score Band Vector Assigner Type

CVSS 3.1 9.6 CRITICAL CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H emo@eclipse.org Secondary published

Version	Score	Band	Vector	Assigner	Type
CVSS 3.1	9.6	CRITICAL	`CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H`	emo@eclipse.org	Secondary	published

Affected Software & Releases

☕ Maven Library Packages

Artifact coordinates and fix boundaries from OSV.dev

Package Coordinate Introduced Fixed Last affected

org.glassfish.main.admingui:admingui 0 8.0.2 unbounded

org.glassfish.jsftemplating:jsftemplating 0 4.2.0 unbounded

Package Coordinate	Introduced	Fixed	Last affected
`org.glassfish.main.admingui:admingui`	0	`8.0.2`	unbounded
`org.glassfish.jsftemplating:jsftemplating`	0	`4.2.0`	unbounded

🖥️ Product CPEs & Version Ranges

3 product(s) over 3 version claims (uncapped)

Source	Vendor	Product	Scheme	Affected Versions
nvd	eclipse	glassfish	generic	＜ 8.0.2
osv	org.glassfish.jsftemplating	jsftemplating	generic	＜ 4.2.0
osv	org.glassfish.main.admingui	admingui	generic	＜ 8.0.2

Source

Vendor

Product

Scheme

Affected Versions

nvd

eclipse

glassfish

generic

＜ 8.0.2

osv

org.glassfish.jsftemplating

jsftemplating

generic

＜ 4.2.0

osv

org.glassfish.main.admingui

admingui

generic

＜ 8.0.2

References

URL	Tags
https://github.com/eclipse-ee4j/glassfish	PACKAGE
https://github.com/eclipse-ee4j/glassfish/releases/tag/8.0.2	WEB
https://gitlab.eclipse.org/security/cve-assignment/-/issues/86	WEB

URL

Timeline

Published 2026-05-19 Last modified 2026-07-24

Published	2026-05-19	By the CVE Program.
NVD record modified	2026-07-24	NVD's own last-modified date for this record.

Published

2026-05-19

By the CVE Program.

NVD record modified

2026-07-24

NVD's own last-modified date for this record.

Pipeline policy audit (15 of 21 rules evaluated)

15 of 21 rules ran. The rest were never reached, because an earlier stage settled the CVE, a distinct claim from a rule that ran and found nothing.

# Rule Outcome Status Detail

identity: Does this identifier describe a vulnerability at all? first match wins

1 rejected no-match CVE Program withdrawal of this identifier

scope: Is the vulnerability ARMR's to address? first match wins

2 armr-patch-file no-match An ARMR patch file on disk for this CVE

3 armr-secure-rule-file no-match An ARMR security-rule template on disk for this CVE

4 manual-classification no-match A researcher's recorded decision for this specific CVE

5 manual classification no-match Manual classification of this CVE in the legacy dataset

6 non-java-upstream no-match A root cause in non-Java code: an OS kernel, a C library, or a bundled runtime

7 oracle-component no-match The defect being in a component of a tracked product the agent does not reach: JavaFX, the Java SE installer, the native WebLogic proxy plug-in. Oracle gives these no CPE of their own and names them only in the description.

8 osv-maven matched Queued for review Maven packages: org.glassfish.main.admingui:admingui, org.glassfish.jsftemplating:jsftemplating

9 cna-maven-package not-reached An affected Maven artifact named by the CNA itself

10 known-cpe-product not-reached An NVD CPE naming a product ARMR tracks

11 cna-cpe-product not-reached A CPE the CNA published naming a tracked product. Oracle publishes these on advisory day, months before NVD.

12 oracle-advisory-family not-reached A tracked product family named in Oracle's own risk matrix. CPE cannot carry a suite name, only a module.

13 non-deployable-product not-reached The only affected product being a sealed appliance image, where Java runs but no -javaagent can be attached

14 unmatched-product not-reached No tracked product matched above, and none present at all

disposition: What do we do about one that is? cumulative: every rule gets a turn

15 patch-hint no-match A reference pointing at the upstream fix: a commit, a PR, a Red Hat bug

16 open-source-maven matched Mitigation candidate Open-source Maven artifact indexed by OSV (Maven packages: org.glassfish.main.admingui:admingui, org.glassfish.jsftemplating:jsftemplating). The source is public, which makes an ARMR patch rule a candidate; nothing here establishes that one can be written.

17 poc-derivable skipped An indexed exploit to reverse-engineer, where there is no source to read

18 no-exploit skipped With no fix path, a public exploit, or a CISA record of one

19 secure-rule-match no-match A vulnerability class already blocked by an ARMR security rule

20 below-action-threshold no-match Severity is CRITICAL — only MEDIUM/LOW are deprioritized

21 no-exploit-published decided it matched No exploit published No working exploit has been published, and an upstream fix path does exist, so the remediation half of the candidacy still holds.

What the scope rules read

The products the record named, and whether known-products.yaml tracks them. This is the evidence the scope stage decided on.

Vendor Product Tracked CPE Source

eclipse glassfish not tracked cpe:2.3:a:eclipse:glassfish:*:*:*:*:*:*:*:*

Raw stored JSON document for this CVE ↗ · NVD ↗ · CVE.org ↗